Chemosphere 53 (2003) 927–934
www.elsevier.com/locate/chemosphere
Competition between alga (Pseudokirchneriella subcapitata),
humic substances and EDTA for Cd and Zn control in
the algal assay procedure (AAP) medium
Celine Gueguen
a,b,*
, Brahim Koukal a, Janusz Dominik
a,b
, Michel Pardos
a
a
b
Institut F.-A. Forel, 10 route de Suisse, CH-1290 Versoix, Switzerland
Centre d’Etudes en Sciences Naturelles de l’Environnement, University of Geneva, 10 route de Suisse, CH-1290 Versoix, Switzerland
Received 21 October 2002; received in revised form 27 June 2003; accepted 15 July 2003
Abstract
The chemical speciation of trace metals in natural waters has important implications for their biogeochemical behavior. Trace metals are present in natural waters as dissolved species and associated with colloids and particles. The
complexation of one trace metal (Cd and Zn at 200 and 390 lg/l respectively) with a green alga Pseudokirchneriella
subcapitata in colloid-free algal culture medium and in presence of colloidal humic substances (HS) is presented. The
inﬂuence of the nature of colloids was also addressed using three ‘‘standard’’ HS: fulvic acid (FA) and, soil (SHA) and
peat humic acids (PHA). The chemical speciation model, MINTEQA2, was used to simulate the inﬂuence of pH and
standardized culture medium on metal association with humic substances. The model was successfully modiﬁed to
consider the diﬀerences in the metal complexation with fulvic (FA) and humic acids (HA). The deviations of concentrations of metals associated with HS between experimental results and model predictions were within a factor of
2. The results of speciation model highlight the inﬂuence of the experimental conditions (pH, EDTA) used for alga
bioassay on the behavior of Cd and Zn. The computed speciation suggests working with a pH buﬀered/EDTA-free
mixture to avoid undesirable competition eﬀects. The behavior of Cd and Zn in solution is more strongly inﬂuenced by
HS than by alga. Metal–HS associations depend on metal and humic substance nature and concentration. Cd is
complexed to a higher extent than Zn, in particular at larger HS concentration, and the complexation strength is in the
order FA < HA.
2003 Elsevier Ltd. All rights reserved.
Keywords: Complexation; Humic substance; Trace metal; MINTEQA2; Ultraﬁltration; Bioassay; Lead; Cadmium
1. Introduction
Knowledge of the chemical forms of metals is essential to understand their interaction with living organisms
*
Corresponding author. Present address: IARC-Frontier,
University of Alaska Fairbanks, 930 Koyukuk Drive, P.O. Box
757335, Fairbanks, AK 99775-7335, USA. Tel.: +1-907-4742642; fax: +1-907-474-2643.
E-mail address: [email protected] (C. Gueguen).
as metal speciation controls their mobility and bioavailability (e.g. Campbell, 1995). Metals in surface water
have traditionally been subdivided into two fractions:
‘‘dissolved’’ and ‘‘particulate’’ according to operationally
deﬁned limit (usually 0.45 lm) and separated by ﬁltration. The ﬁlter passing ‘‘dissolved’’ fraction clearly does
not represent the truly dissolved metal ions, but is composed of free metals, metals bound to a variety of ligands,
forming molecules of various dimensions and chemical
characteristics, which may further be bound to larger
entities of colloidal size (Buﬄe and van Leeuwen, 1992).
0045-6535/$ - see front matter 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0045-6535(03)00719-7
928
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
Assessing toxicity of contaminated waters with algal
bioassays is one of the standard tools in environmental
impact assessment. It has however been noticed that at
similar metal content bioassays may show a wide range
of toxic eﬀects. Evidence exist that the total metal concentration is not a good predictor of bioavailability,
toxicity and mobility of the metal (Sunda and Guillard,
1976; Sunda, 1991). The speciation of a metal is expected
to greatly aﬀect the availability of trace elements for
phytoplankton and other organisms. With regards to
their reactivity, metal can be associated with particles
(e.g. alga), colloids (e.g. humic substances, HS) and
dissolved species (e.g. EDTA). Free metal ions are
thought to be the most available and toxic (e.g. Campbell, 1995 and therein references). In natural waters,
only a small fraction of dissolved metal may be present
as free hydrated cations because of the presence of a
large variety of inorganic and organic ligands. Humic
substances, which are one of the main constituents of the
natural aquatic colloids (Raspor, 1980; Linnik, 2000;
Gueguen and Dominik, 2003), play a dominant role in
the metal ion distribution and aﬀect the metal bioavailability. The eﬀects of metal speciation on toxic effects are addressed in the companion paper (Koukal
et al., this volume).
The metal partitioning between particulate, colloidal
and truly dissolved fractions was examined in solutions
constituted by a trace metal (Me: Cd, Zn), humic substance (HS), alga and standardized culture medium
(AAP medium; USEPA, 1994). The presence of EDTA
in AAP medium, the eﬀects of pH and the inﬂuence of
nature and contents of HS on metal complexation were
also studied by chemical speciation modeling. The
MINTEQA2 model is a speciation model containing a
default database on metal–DOM complexation and was
extensively used in aqueous systems (e.g., Christensen
and Christensen, 1999; Kocaoba and Akcin, 2002). In
order to evaluate the role of alga and HS in metal
complexation, the distribution coeﬃcients between particulate and truly dissolved, colloidal and truly dissolved, and particulate and total dissolved fractions
were calculated as Kp , Kc , Kd respectively. Laboratoryderived Kp , Kc , Kd values have advantages over those
determined in natural environments in that individual
species can be singled out. Thus there is control over the
algal species and the organic matter quality.
2. Materials and methods
2.1. Preparation of solutions
Experiments were performed on three diﬀerent humic
substances (HS) types purchased from International
Humic Acid Society (IHAS). Their nature, origin and
molecular weight are diﬀerent: (1) fulvic acid (FA) from
Suwannee River has smaller molecular weight (103
kDa, Aiken et al., 1985) than (2) humic acid from soil
(SHA) and (3) humic acid from peat (PHA) which has
the highest molecular weight (104 –105 , Aiken et al.,
1985). Detailed composition of HS has been reported in
the companion paper (Koukal et al., this volume). All
HS stock solutions were daily prepared at 250 mg/l with
ultrapure water (Millipore system, >18 MX). They were
diluted to a ﬁnal HS concentrations of 1 and 5 mg HS/l.
For each HS contents, one metal (Cd or Zn) was added.
Cd and Zn were added to a ﬁnal concentration of 200
and 390 lg/l respectively. Although these concentrations
are not representative of the levels encountered in natural aquatic environments, except of extremely polluted
streams, they correspond to the EC50 calculated for the
studied green alga in a short-term (1 h) bioassay (Koukal et al., this volume). The EC50 means the eﬀective
concentration of metal needed to induce the 50% inhibition of the [14 C] assimilation by algae.
The experiments were made in the algal assay procedure medium (AAP) (USEPA, 1994). In addition to
nutrients, iron and EDTA used keep iron in solution
(Lewin and Chen, 1971; Anderson and Morel, 1982;
Hughes and Poole, 1991) are present. The detailed
chemical composition of AAP medium is described in
Koukal et al. (this volume).
Each solution (Me + HS) was stored in the dark at
ambient temperature. After 24 h equilibration time, the
stock culture of the green alga, Pseudokirchneriella
subcapitata (formerly named Selenastrum capricornutum) was directly inoculated in the solution. The culture
was harvested at a cell density of 105 cells/ml and then
incubated during 1 h at 24 C under 5000–6000 Lux light
(USEPA, 1994). All experiments were made in algal
medium (AAP medium; USEPA, 1994) at pH ¼ 8.5 0.1 using 1% of CH3 COOH/CH3 COONH4 mixture
(2 mmol/l).
2.2. Filtration/ultraﬁltration
After incubation, the mixture (AAP + Me + HS + alga)
was ﬁltered with an acid-washed 1 lm polycarbonate
Nuclepore ﬁlter to remove the particles (here constituted
by algae). The ﬁlter was dried in acid-washed plastic
boxes and the weight of material retained on ﬁlter determined. The ﬁltered solution was further ultraﬁltered
through a 1 kDa regenerated cellulose cartridge (Prepscale, Millipore) to isolate colloidal from the truly dissolved fraction. The ultraﬁltration protocol has been
described previously (Gueguen et al., 2002). Brieﬂy, the
ultraﬁltration cartridge was thoroughly cleaned with
0.5 N superpure HCl, ultrapure water until pH 6 is
reached in the permeate and retentate, and 0.1 N superpure NaOH. Finally, the system was ﬂushed with a
large volume of Milli-Q water until the permeate and
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
retentate were free of noticeable residual organic carbon.
Before ultraﬁltration, the cartridge was preconditioning
with the sample. The applied concentration factor for
ultraﬁltration process was about 2 for all solutions.
Bulk, ﬁltered, colloidal and truly dissolved fractions
were collected and acidiﬁed to pH < 2 with superpure
HNO3 to prevent the metals from adsorbing to the
container material or precipitating. They were stored in
darkness at 4 C until metal analysis.
2.3. Measurements
pH and conductivity were measured before and after
equilibration time. The pH after equilibration did not
change by more than 0.05 pH-unit and conductivity
increased from 340 to 500 lS/cm for all experiments.
Concentrations of organic carbon (Corg) were determined by absorbance measurements at 254 nm (Spectronic 1201, Milton Roy Company) immediately after
sampling. A calibration curve for the absorbance of
fulvic acid, peat and soil humic acid solutions was prepared by diluting the appropriate stock solution to the
concentration range encountered in the colloidal and
truly dissolved fractions.
The samples were mineralized in HNO3 /H2 O2 mixture prior to metal analysis by inductively coupled
plasma mass spectrometry (ICPMS) (Agilent HP4500).
Rhodium was used as an internal standard for all
measurements. The accuracy of the metal determinations was checked on a regular basis using the 1643d
standard (National Institute of Standards and Technology).
Mass balance of the ultraﬁltration process was acceptable for Cd (89 4%, n ¼ 7), Zn (95 5%, n ¼ 12)
and Corg (101 4%, n ¼ 19).
929
complex material consisting of many diﬀerent types of
monoprotic acid sites. This is a composite ligand model
with a Gaussian aﬃnity distribution (Dobbs et al., 1989;
Susetyo et al., 1991; Allison and Perdue, 1994). The
electrostatic interactions are not taken into account explicitly. The concentration of the binding sites is normally distributed with respect to their log K values for
proton or metal binding. A database available for proton and metal interaction with Suwannee River DOM
and their mean log K values is included (i.e. log K ¼ 3:3
and 3.5 for Cd–DOM and Zn–DOM respectively). The
composite ligand component DOM represents a complex mixture of ligands without distinction between
humic and fulvic fractions.
Metal binding to humic substances can occur at two
diﬀerent types of binding sites, type A site (associated
with carboxylic groups) and type B (associated with
phenolic type group) (Buﬄe, 1988 and references
therein). Despite the binding sites are represented only
by one type of site (carboxylic) in MINTEQA2 model,
the predictions of metal complexation by DOM can be
reasonable (Christensen and Christensen, 1999).
3. Results and discussion
Table 1 shows the partitioning of organic carbon
between colloidal (>1 kDa) and truly dissolved (<1 kDa)
fractions. Although fulvic acids have the molecular
weight close to the nominal cut-oﬀ of the membrane,
organic carbon was nearly totally found in the colloidal
fraction (>92%). This suggests a considerable coagulation of fulvic acids in the AAP medium. As expected,
humic acids were almost entirely retained in the colloidal
fraction.
2.4. Chemical speciation model
MINTEQA2 (Allison et al., 1991) is a speciation
model extensively used for calculating the inorganic
aqueous species. In the version 3.11, it includes a submodel for computing the complexation of various metal
cations with dissolved organic matter (DOM) (Dobbs
et al., 1989; Allison et al., 1991; Allison and Perdue,
1994; Serkiz et al., 1996). This submodel is based on the
work of Dobbs et al. (1989) who considered DOM as a
4. Simulation by the computer speciation model: MINTEQA2
4.1. Inﬂuence of pH
Complexation of metal with HS liberates protons
inducing pH modiﬁcation in a closed laboratory system.
To evaluate the pH impact on metal speciation, a diagram showing the relative proportion of each metallic
Table 1
Repartition (%) of organic carbon between colloidal and truly dissolved after ultraﬁltration of studied solutions containing AAP and 1
or 5 mg/l of HS
FA (river)
Colloids
Truly dissolved
SHA (soil)
PHA (peat)
1 mg/l
5 mg/l
1 mg/l
5 mg/l
1 mg/l
5 mg/l
92
8
98
2
95
5
95
5
99
1
95
5
930
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
100
100
HS 1mg/l
HS 1mg/l
75
Zn (%)
Cd (%)
75
50
50
25
25
0
0
6
7
pH
8
6
9
7
8
100
100
HS 5mg/l
HS 5mg/l
75
Zn (%)
75
Cd (%)
9
pH
50
50
25
25
0
0
6
7
8
pH
9
6
7
8
9
pH
Fig. 1. Speciation of Cd(II) and Zn(II) in modiﬁed AAP (model calculation). (}) Me2þ , (j) Me–HS, (N) MeCO3 , (M) Me(CO3 )2
2 .
species in water as a function of pH is developed (Fig. 1).
According to the MINTEQA2 calculations, free metal
ion is the dominant form at pH < 7 and 1 mg/l HS for
both metals. The calculation shows the tendency of
CO2
to form complex in solution with metal in the
3
natural pH range (7 < pH < 9). Thus the change of pH
in the range from 6 to 9 may strongly inﬂuence metal
complexation. The use of buﬀer maintaining the pH at
8.5 appears necessary. Such pH is desirable to simulate conditions typical for lake epilimnion during a
high productivity period. Ammonium acetate buﬀer was
chosen for several reasons: it does not disturb signiﬁcantly organic carbon measurements, it is not toxic for
alga at concentration used it does not contain a noticeable amount of metals (<10 ng/l). In addition, our
preliminary modeling showed that it did not constitute a
signiﬁcant ligand for studied metals.
4.2. Inﬂuence of culture medium AAP
The standard culture medium contains high concentration of EDTA (0.8 lM), a strong ligand, and Fe3þ
(0.59 lM) that is maintained in solution by EDTA. The
model simulation of two Me + HS solutions with standard AAP and modiﬁed AAP (without EDTA and Fe3þ )
was performed at pH ¼ 8.5 (Table 2). The model suggests that Me–HS complexation is notably inﬂuenced by
the presence of EDTA especially for cadmium. The
predictions of the eﬀects of metal complexation by HS in
normal AAP were lower than in EDTA-free AAP (1.7–
5.6% and 13.2–45.0% for Cd, respectively). HS and
EDTA can be therefore in competition for metal complexation. Moreover, the predicted proportion of free
ionic Cd2þ would be less important in the presence of
EDTA (10%) than in EDTA-free medium (80%). The
competition with EDTA has minor inﬂuence on formation of Zn–HS complex and on the proportion of
Zn2þ .
EDTA in a standard AAP medium recommended by
USEPA (1994) is necessary in long-term tests to maintain Fe in solution, an important oligoelement for alga.
The presence of EDTA however inﬂuences the proportion of Me associated with HS and decreases the free
metal ion contents (particularly for Cd). Thus, the
‘‘true’’ toxicity of sample may be underestimated using
the standard AAP.
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
Table 2
Formation of Me–DOM complexes (%) in standard and modiﬁed (EDTA-free) culture medium AAP at 1 and 5 mg/l HS
(model calculation, pH ¼ 8.5). The experimental data (n ¼ 9)
(in parenthesis) was averaged over all HS types
Normal AAP
EDTA-free AAP
1 mg/l
HS
5 mg/l
HS
1 mg/l
HS
5 mg/l
HS
Zn2þ
Zn–EDTA
Zn–DOM
Zn(OH)2
Zn(OH)þ
26
16
1
44
10
20
16
23
32
8
31
24
6 (29 5)
40
10
26 (53 12)
39
9
Cd2þ
Cd–EDTA
Cd–DOM
10
87
2
6
87
6
80
51
13 (25 11)
45 (54 23)
Log K (Me–EDTA) ¼ 16.27 and 16.53 for Cd and Zn respectively.
Modeling calculations show that culture medium
constituents and pH inﬂuence Me–HS formation. Consequently, a modiﬁed AAP and the use of buﬀer are
recommended to study the Me–HS interactions during
the short-term (1-h test) biotests, especially for Cd. It
should, however, be noticed that providing suﬃcient
amount of Fe is essential in long-term test, which is not
assured without the presence of EDTA.
4.3. Eﬀects of HS complexation: comparison of experimental results and model predictions
• Metal contents.
The total concentrations of Cd and Zn used in this
study were chosen at the EC50 level for the studied alga
after 1 h exposure. They were 200 and 390 lg/l for Cd
and Zn respectively (Koukal et al., this volume). It can
be noticed that these high concentration may not be
relevant for natural aquatic environments. Speciation
calculation was performed to evaluate the impact of HS
on the behavior of each metal at the same metal contents. The increase of Cd concentration from 200 to 390
lg/l did not inﬂuence its complexation with HS (data not
931
shown). Thus the diﬀerences observed in the experiments
were not due to metal concentration eﬀects.
• Content of HS.
Eﬀects of raising the HS contents from 1 to 5 mg/l on
metal sorption, while maintaining a constant metal
concentration are illustrated in Fig. 2. As the HS concentration increased, a larger proportion of metals were
complexed by the HS. The Me–HS complexation increased by a factor of about 2 for the three studied HS
for both test metals (Table 2). At the same time, the
metal contents in the truly dissolved fraction decreased.
The results obtained by modeling generally underestimated the HS complexed fraction as compared to the
experimental data. The deviations of the metal associated with HS between experimental results and model
predictions were within a factor of 2. Considering the
uncertainties usually associated with thermodynamic
data on complexation reactions, the experimental data
and model predictions were not unacceptable (Christensen and Christensen, 1999).
• Nature of HS.
At the same concentration of organic colloids, a
larger fraction of Cd was complexed with HA than with
FA (Fig. 2). This diﬀerence is less pronounced for Zn.
The proportion of Zn associated with FA was larger
than those of Cd. That can be explained by the abundance of carboxylic functional groups in FA favoring
the complexation of type-A metal such as Zn over a
type-B metal, Cd. For PHA, especially at the concentration of 5 mg/l, Cd is complexed in a much larger
proportion than Zn. This may be explained by preferences of a type-B metal for nitrogen ligands.
4.4. Modiﬁcation of the MINTEQA2-model
In the MINTEQA2-model, there is no distinction
between HA and FA. However, the experimental data
showed that trace metals were more or less complexed
depending on the nature of HS. The model should
therefore be modiﬁed to take into account these diﬀerences. Speciﬁc metal binding properties for HS were
derived from the literature (Buﬄe, 1988; Allison and
Perdue, 1994; Tipping, 1994) under the assumption that
80
Cd
Zn
% Me-HS
60
40
20
0
FA
SHA
PHA
FA
SHA
PHA
Fig. 2. Percent of metal associated with FA, SHA and PHA at 1 mg/l HS (black) and 5 mg/l HS (white) in the dissolved fraction,
determined with 1 kDa membrane ultraﬁltration.
932
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
ments, 9 3% of metal were associated with algal cells.
No signiﬁcant competition between alga and HS were
noted for metal sorption.
80
line 1:1
Modeling (% Me-HS)
60
4.6. Environmental implications
2
3
1
As shown above, the association Me–HS depends on
the nature and the concentration of HS, as well as on the
metal. The distribution coeﬃcients (Kd , Kp and Kc ) for
metals associated with alga, colloids (here HS) and
present as truly dissolved species (aqua ions, small inorganic and organic complexes) can be calculated as
follows:
40
2
1
20
3
2
1
1
3
23
0
0
20
40
60
80
Experiment (% Me-HS)
Fig. 3. Experimental results compared with model predictions
for () Cd and ( ) Zn associated with 1 FA, 2 SHA, 3 PHA at 1
mg/l (ﬁlled) and 5 mg/l (open). Modelling parameters: Ka : 3.3
and 4.82 for FA and HA respectively; site density: 16.6 and 13
meq/g for FA and HA respectively.
SHA and PHA have the same complexing capacity and
aﬃnity for metals as HA (see legend Fig. 3). Ka and site
density of HS are the parameters whose values have a
signiﬁcant impact on the results.
The model gave excellent prediction of metal complexation to FA (Fig. 3) but slightly underestimated the
complexation with SHA and PHA. On the other hand,
the complexation increase due to the changes in HS
content was well reﬂected by model predictions. However, more studies are needed to apply this model to
natural aquatic environments where lower metal concentrations are expected.
4.5. Metal associated with alga
The type and contents of HS did not inﬂuence highly
the metal associated with alga (Fig. 4). In all experi-
Kd ðl=kgÞ ¼
mass of particulate metal=mass solids
mass of total dissolved metal=volume of water
Kp ðl=kgÞ ¼
mass of particulate metal=mass solids
mass of truly dissolved metal=volume of water
Kc ðl=kgÞ ¼
mass of colloidal metal=mass colloids
mass of truly dissolved metal=volume of water
They characterize the metal aﬃnity for each phase.
While the distribution coeﬃcients between particulate and total dissolved fractions (<1.2 lm) have been
extensively studied (e.g. Radovanovic and Koelmans,
1998; Gueguen et al., 2000; Paulson and Gendron,
2001), the data on the partitioning between particles and
colloids, and colloids and truly dissolved (<1 kDa) are
still scarce (Admiraal et al., 1995; Sa~
nudo-Wilhelmy
et al., 1996; Wen et al., 1999; Gueguen and Dominik,
2003). The partitioning coeﬃcients calculated from experimental data (Table 3) give some information on the
role of HS in metal complexation.
For each metal, the values of Kp are larger than those
of Kd , which indicates the important role of HS during
% Cd associated with algae
25
Cd
Zn
20
15
10
5
0
FA
SHA
PHA
FA
SHA
PHA
Fig. 4. Percent of metal associated with alga in presence of FA, SHA and PHA at 1 mg/l (black) and 5 mg/l (white). Except for Cd with
FA, the % of metal associated with alga were not signiﬁcantly diﬀerent when HS contents increased.
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
933
Table 3
Distribution coeﬃcients (l/kg) determined in the studied solutions (modiﬁed AAP + alga + HS + Me)
[HS] (mg/l
Kd (104 l/kg)
Kp (104 l/kg)
Kc (104 l/kg)
Mean
3r
Mean
Mean
3r
Cd + FA
Cd + PHA
Cd + SHA
Zn + FA
Zn + PHA
Zn + SHA
1
1
1
1
1
1
2.3
3.5
2.4
2.6
2.0
1.7
0.3
0.3
0.2
0.1
0.1
0.3
3.0
5.7
4.0
3.6
2.8
3.0
0.4
0.4
0.3
0.2
0.2
0.5
23.0
47.0
26.0
48.0
54.0
59.0
0.5
0.9
0.2
0.6
0.5
15.0
Cd + FA
Cd + PHA
Cd + SHA
Zn + FA
Zn + PHA
Zn + SHA
5
5
5
5
5
5
7.2
3.3
4.7
2.7
4.0
1.3
0.3
0.6
0.3
0.1
0.3
0.2
12.0
16.0
14.0
5.1
8.4
4.2
0.5
28.0
0.9
0.2
0.5
0.5
15.0
80.0
41.0
30.0
23.0
44.0
0.4
1.1
0.3
0.2
0.1
0.8
the metal partitioning. The values of Kc are much higher
than those of Kp , which means that the complexation
capacity of HS (colloids) is greater than that of the alga
(particles) (Wen, 1996).
The partition coeﬃcients Kd calculated in this study
were compared with those found in natural environments during an algal bloom. In our experiments,
the average Kd was 3.9 1.9 104 l/kg for Cd and
2.4 0.9104 l/kg for Zn. In a phytoplanktonic bloom,
partitioning coeﬃcient Kd of Cd and Zn were 8.3 104
and 7.9 104 l/kg respectively, in the Rhine River (Admiraal et al., 1995) and 0.7–5.1 104 and 0.1–3.8 104
l/kg respectively in the Lake Geneva (Gueguen, 2001).
Despite the high contents of trace metals and diﬀerent
matrix used in the lab experiments, a similar partitioning
between particle/water was found in the lab and in situ
measurements.
5. Conclusion
The calculation of chemical speciation modeling
performed on MINTEQA2 showed that to study metal
complexation in the presence of alga and humic substances (HS), the algal culture medium must be modiﬁed
to not underestimate the Me–HS formation and that the
use of a buﬀer is needed to avoid the pH ﬂuctuation
occurring during the equilibration time, aﬀecting the HS
complexation. Contrary to the complexation with alga,
metal contents and nature of HS inﬂuenced the association Me–HS. These modeling calculations are in fairly
good agreement with the experimental results. The distinction of humic acids and fulvic acids in the MINTEQA2-model was performed in this study and showed a
relatively good agreement with the experimental data. It
has been shown that humic substances are much more
important ligand for Cd and Zn than alga.
3r
Acknowledgements
This research was funded by the Swiss National
Science Foundation (‘‘Partition of trace metals and their
bioavailability in continental surface waters’’ grant no.
20-57189.99). Critical comments of two anonymous reviewers are acknowledged.
References
Admiraal, W., Tubbing, G.M.J., Breebaart, L., 1995. Eﬀects of
phytoplankton on metal partitioning in the lower river
Rhine. Wat. Res. 29, 941–946.
Aiken, G.R., McKnight, D.M., MacCarthy, P., 1985. Humic
Substances in Soil, Sediment and Water: Geochemistry,
Isolation and Characterization. Wiley Interscience, New
York. 692pp.
Allison, J.D., Brown, D.S., Novo-Gradac, K.J., 1991. MINTEQA2/PRODEFA2, A geochemical assessment model for
environmental systems: Version 3.0 Users manual. EPA/
600/3-91/021. US Environmental Protection Agency, Athens, GA.
Allison, J.D., Perdue, E.M., 1994. Modelling metal–humic
interactions with MINTEQA2. In: Senesi, N., Miano, T.M.
(Eds.), Humic Substances in the Global Environment and
Implications on Human Health. Elsevier Science, Amsterdam, pp. 927–942.
Anderson, M.A., Morel, F.M.M., 1982. The inﬂuence of
aqueous iron chemistry on the uptake of iron by the coastal
diatom Thalassiosira weissﬂogi. Limnol. Oceanogr. 27, 789–
813.
Buﬄe, J., 1988. Complexation Reactions in Aqueous Systems:
An Analytical Approach. Wiley & Sons Publishers, Chichester, England. 692pp.
Buﬄe, J., van Leeuwen, H.P., 1992. Environmental particles,
vol. 1. Lewis publishers, Chelsea, Michigan, 554pp.
Campbell, P.G.C., 1995. Interaction between trace metals and
aquatic organisms: A critique of the free-ion activity model.
In: Tessier, A., Turner, D.R. (Eds.), Metal Speciation and
934
C. Gueguen et al. / Chemosphere 53 (2003) 927–934
Bioavailability in Aquatic Systems. John Wiley & Sons
LTD., Chichester, England, pp. 45–102.
Christensen, J.B., Christensen, T.H., 1999. Complexation of
Cd, Ni and Zn by DOC in polluted groundwater: A
comparison of approaches using resin exchange, aquifer
material sorption, and computer speciation models
(WHAM and MINTEQA2). Environ. Sci. Technol. 33,
3857–3863.
Dobbs, J.C., Susetyo, L.A., Knight, F.E., Carreira, L.A.,
Azarraga, L.V., 1989. Characterization of metal binding
sites in fulvic acids by lanthanide ion probe spectroscopy.
Anal. Chem. 61, 483–488.
Gueguen, C., 2001. Role of colloid in trace metal partition in
contrasted aquatic environments. Ph.D. thesis, University of
Geneva, 286pp.
Gueguen, C., Belin, C., Dominik, J., 2002. Organic colloid
separation in contrasting aquatic environments with tangential ﬂow ﬁltration. Wat. Res. 36, 1677–1684.
Gueguen, C., Dominik, J., 2003. Partitioning of trace metals
between particulate, colloidal and truly dissolved fractions
in a polluted river: The Upper Vistula River (Poland). Appl.
Geochem. 18, 457–470.
Gueguen, C., Dominik, J., Pardos, M., Benninghoﬀ, C.,
Thomas, R.L., 2000. Partition of metals in the Vistula
River and in eﬄuents form sewage treatment plants in the
region of Cracow (Poland). Lakes Reservoirs: Res. Manage.
5, 59–66.
Hughes, M.N., Poole, R.K., 1991. Metal speciation and
microbial growth––the hard (and soft) facts. J. General
Microbiol. 137, 725–734.
Kocaoba, S., Akcin, G., 2002. Removal and recovery of
chromium and chromium speciation with MINTEQA2.
Talanta 57, 23–30.
Koukal, B., Gueguen, C., Pardos, M., Dominik, J., this volume.
Eﬀect of humic substances on Cd and Zn toxicity to
Pseudokirchneriella subcapitata.
Lewin, J., Chen, C.H., 1971. Available iron: A limiting factor
for marine phytoplankton. Limnol. Oceanogr. 16, 670–675.
Linnik, P.M., 2000. Zinc, lead and cadmium speciation in
Dnieper water-bodies. Lakes Reservoirs: Res. Manage. 5,
261–270.
Paulson, A.J., Gendron, J.F., 2001. Partitioning of copper at
concentrations below the marine water quality criteria.
Environ. Tox. Chem. 20, 952–959.
Radovanovic, H., Koelmans, A.A., 1998. Prediction of in situ
trace metal distribution coeﬃcients for suspended solids in
natural waters. Environ. Sci. Technol. 32, 753–759.
Raspor, B., 1980. Distribution and speciation of cadmium in
natural waters. Part I: Ecological cycling. In: Nriagu, J.O.
(Ed.), Cadmium in the Environment. Wiley & Sons, New
York, pp. 147–236.
Sa~
nudo-Wilhelmy, S.A., Rivera-Duarte, I., Flegal, A.R., 1996.
Distribution of colloidal trace metals in the San Francisco
Bay estuary. Geochim. Cosmochim. Acta 60, 4933–4
944.
Serkiz, S.M., Allison, J.D., Perdue, E.M., Allen, H.E., Brown,
D.S., 1996. Correcting errors in the thermodynamic database for the equilibrium speciation model MINTEQA2.
Wat. Res. 30, 1930–1933.
Sunda, W.G., 1991. Trace metal interactions with marine
phytoplankton. Biol. Oceanogr. 6, 411–441.
Sunda, W.G., Guillard, R.R.L., 1976. The relationships
between cupric ion activity and the toxicity of copper to
phytoplankton. J. Marine Res. 34, 511–529.
Susetyo, W., Carreira, L.A., Azarraga, L.V., Grimm, D.M.,
1991. Fluorescence techniques for metal–humic interactions.
Fresenius J. Anal. Chem. 339, 624–635.
Tipping, E., 1994. WHAM––a chemical model and computer
code for waters, sediments and soil incorporating a discrete
site/electrostatic model of ion-binding by humic substances.
Comput. Geosci. 20, 973–1023.
USEPA, 1994. Short term methods for estimating chronic
toxicity of eﬄuents and receiving waters to freshwater
organisms. EPA/600/4-91/002. US Environmental Protection Agency, Cincinnati, Ohio.
Wen, L.S., 1996. Geochemistry of trace metals in estuarine
waters: A multi-phase speciation approach. Ph.D. thesis,
Texas A&M University, 217pp.
Wen, L.S., Santschi, P.H., Gill, G., Paternostro, C., 1999.
Estuarine metal distribution in Galveston Bay. I: Colloidal
forms and dissolved phase speciation. Mar. Chem. 63, 185–
212.